PSIHOLOGIJA, 2014, Vol. 47(4), 373–391
© 2014 by the Serbian Psychological Association
UDC 159.923.2/.3.075
DOI: 10.2298/PSI1404373B
Validation of two conceptualizations of fragile
self-esteem: contingent high self-esteem and
incongruent high self-esteem
Bojana Bodroža
Institute for Educational Research, Belgrade, Serbia
The aim of this research was to validate two aspects of fragile high self-esteem: a
combination of contingent and high (explicit) self-esteem and a combination of high explicit
and low implicit self-esteem (i.e. incongruent high self-esteem), as well as to examine the
relationship between these aspects of fragile self-esteem and narcissism. No convergence was
found between contingent high and incongruent high self-esteem. The result was consistent
regardless of the technique of measurement of implicit self-esteem. There was a limited
evidence that individuals with narcissistic personality characteristics were characterized by
high self-esteem contingent upon competences, but not by a combination of high explicit and
low implicit self-esteem, as an aspect of fragile self-esteem. Also, individuals with low selfesteem more contingent upon competences showed higher levels of narcissistic characteristics
than those who were not contingent in this domain.
Keywords: fragile self-esteem, implicit self-esteem, contingent self-esteem, Implicit
Association Test, Name Letter Preference, narcissism
Over the last two decades, researchers in the field of self-esteem and
identity have focused their attention not only on levels of self-esteem, but also
on the quality of self-esteem. Early research in this field showed that high selfesteem (HSE) is usually more beneficial and adaptive than low self-esteem
(LSE). As compared to people with HSE, people with LSE are more prone to
Corresponding author: [email protected]
Note. This article is the result of the projects “From encouraging initiative, cooperation and
creativity in education to new roles and identities in society“ (No. 179034) and “Improving
the quality and accessibility of education in modernization processes in Serbia” (No. 47008),
financially supported by the Ministry of Education, Science, and Techological Development
of the Republic of Serbia (2011–2015).
Acknowledgement: I would like to thank professors Goran Opačić, PhD, Petar Čolović, PhD,
and Petar Milin, PhD, for their invaluable help in organizing and conducting this research. I
am also very grateful to Janko Međedović and Boban Petrović for giving me useful advice
and comments on the paper, and to Aleksandra Savić for proofreading the paper.
374
VALIDATION OF TWO CONCEPTUALIZATIONS OF FRAGILE SELF-ESTEEM:
CONTINGENT HIGH SELF-ESTEEM AND INCONGRUENT HIGH SELF-ESTEEM
depression (Tennen & Affleck, 1993), anxiety (Battle, Jarratt, Smit, & Precht,
1988; Rawson, 1992), social anxiety (Leary & Kowalski, 1995), loneliness (Vaux,
1988), general variability in emotional states (Campbell, Chew, & Scratchley,
1991), more susceptible to persuasion (Janis, 1954), and generally less satisfied
with their life (Diener, 1984; Myers & Diener, 1995). In the seventies HSE was
promoted as a kind of “social vaccine”, which could prevent a wide range of
problems, from drug abuse to juvenile pregnancy (The California Task Force
to Promote Self-Esteem and Personal and Social Responsibility, 1990, as in
Baumeister, 1998, p. 698) and, therefore, it was highly recommended to parents
to foster HSE in their children.
HSE is usually accompanied by self-enhancement tendencies including
the “better-than-average” effect – an overestimation of one’s positive qualities
and underestimation of one’s negative qualities relative to others (Brown, 2012)
and defensive reactions to ego threats. These cognitions and behaviors were first
considered adaptive, because they protect individuals from fluctuations in the
sense of self-worth caused by daily negative events (Taylor & Brown, 1988).
Nevertheless, research that followed suggested not only that HSE individuals
could show maladaptive behaviors (e.g. Baumeister, Smart, & Boden, 1996;
McFarlin, Baumeister, & Blascovich, 1984), but also that positive illusions
and self-enhancement tendencies were not so beneficial in the long run as was
initially thought. Research indicated that, over a longer period, self-enhancement
leads to a decrease in self-esteem and well-being, as well as to increased
disengagement from academic context (Robins & Beer, 2001). It is also related
to the deteriorating pattern of interpersonal perceptions (Paulhus, 1998), poor
social skills, and psychological maladjustment (Colvin, Block, & Funder, 1995).
From that point on, psychologists started looking for the ways to differentiate
between “good” and “bad” forms of HSE.
Secure and fragile self-esteem
“Secure” (also true, genuine, non-defensive, optimal) high self-esteem
represents a benign form of high self-esteem. Persons with secure HSE have
a solid and realistic base for their feeling of self-worth and are ready to accept
themselves with all their imperfections (Zeigler-Hill, Besser, & King, 2011).
They do not react defensively after personal failures, their estimate of their selfworth is not dependent on others’ liking, living up to others’ expectations or
meeting certain standards of excellence (Crocker & Wolfe, 2001; Johnson &
Blom, 2007; Park & Crocker, 2008), and they do not need to be superior to
others (Kernis, 2003).
Contrary to the secure form of HSE, there is a malign form of HSE called
“fragile” (also defensive). It involves “positive self-feelings that are vulnerable
to challenge and require continual promotion and protection” (Kernis et al.,
2005, p. 312). Behavioral and cognitive manifestations of fragile self-esteem
include vulnerability of self-image to threats to self-esteem, a constant need for
confirmation and validation of one’s self-worth, and reliance of self-image upon
Bojana Bodroža
375
some degree of self-deception (Zeigler-Hill et al., 2011). Individuals with fragile
self-esteem act defensively after failure or any kind of ego-threat, have a strong
need for self-promotion, and show other dysfunctional behavioral patterns
(Kernis & Paradise, 2002).
Kernis lists three aspects of HSE that help distinguish its secure and fragile
forms: contingency, congruence, and stability of high self-esteem (Kernis, Lakey,
& Heppner, 2008; Kernis & Paradise, 2002). Contingency of self-esteem refers
to feelings of self-worth which are dependent on continual validation, meeting
certain standards of excellence (or even being perfect) and/or on living up to
others’ expectations (Crocker, 2002; Kernis et al., 2005; Kernis & Paradise,
2002). Thus, contingency characterizes fragile HSE, whereas secure self-esteem,
which does not depend on the attainment of specific outcomes or the acceptance
of others, is non-contingent.
Congruency of self-esteem refers to a correspondence between conscious
and pre– or unconscious beliefs and feelings toward the self. Explicit self-esteem
(ESE) is usually defined as a conscious feeling of acceptance, self-worth and
self-liking (Zeigler-Hill, 2006) and it is measured by self-report questionnaires
(e.g. Rosenberg Self-Esteem Scale; Rosenberg, 1965). On the other hand,
implicit self-esteem (ISE) is believed to consist of preconscious, automatic,
and overlearned self-evaluations (Greenwald & Banaji, 1995). Since ISE is
inaccessible through conscious cognitive processes, researchers have developed
special techniques by which to access preconscious aspects of self-regard. The
most frequently used instruments for measuring ISE are Self-Esteem – Implicit
Association Test (SE-IAT) (Greenwald & Farnham, 2000; Greenwald, McGhee,
& Schwartz, 1998) and Name Letter Preference test (NLP) (Koole, Dijksterhuis,
& van Knippenberg, 2001; Nuttin Jr, 1985, 1987). When determining congruency
between ESE and ISE, researchers take ESE as a reference point. A combination
of high ESE and low ISE is called incongruent or discrepant high self-esteem
and, as such, it is considered to be a manifestation of fragile SE. A combination
of high ESE and high ISE is referred to as congruent self-esteem and represents
the secure form of self-esteem.
The third aspect of fragile and secure HSE takes into account stability of
self-esteem. (In)stability of self-esteem refers to the feeling of self-worth that
fluctuates across time and situations (Kernis, Cornell, Sun, Berry, & Harlow,
1993). The greater the magnitude of fluctuations of self-esteem, the more
unstable it is. High self-esteem that fluctuates often and to a greater extent is a
manifestation of fragile self-esteem, while stable HSE is a manifestation of the
secure form of HSE.
Convergent validity of three aspects of fragile self-esteem
Contingent, incongruent, and unstable high self-esteem, although all seen
as aspects of fragile HSE, are not considered to be the same or interchangeable
constructs. Thus, we would expect these constructs to correlate to a moderate
degree. However, not many studies have addressed the question of convergent
376
VALIDATION OF TWO CONCEPTUALIZATIONS OF FRAGILE SELF-ESTEEM:
CONTINGENT HIGH SELF-ESTEEM AND INCONGRUENT HIGH SELF-ESTEEM
validity of the three aspects of fragile self-esteem. The few studies exploring
this question have shown that there are significant but quite modest correlations
between contingent and unstable self-esteem (Crocker & Wolfe, 2001; Kernis et
al., 1993; Kernis & Paradise, 2002).
Kernis and his colleagues (Kernis et al., 2008) examined the correlation
between aspects of contingent, implicit and stable SE. They concluded that
contingent and instable SE correlate positively, while they both correlate
negatively with ISE measured by NLP. In other words, the lower the implicit
self-esteem, the higher its contingency and instability. Also, moderate to low
correlations between the three suggest that these constructs are not completely
redundant. The problem with the results obtained in this research is that Kernis
and his colleagues did not compare combinations of contingency or stability
with the level of self-esteem (i.e. contingent HSE and instable HSE), or the
combination of level of explicit and implicit SE, which means that they did not
directly validate the construct of fragile self-esteem.
The proof of convergent validity of three aspects of fragile and secure
SE usually comes from research on their behavioral correlates, i.e. defensive
behaviors after threats to self-esteem. Their similarity lies in the fact that they all
have outcomes relevant to the feeling of self-worth.
HSE individuals that base their feeling of self-worth on the acceptance
of others, respond to negative feedback with more self-presentational concerns
than HSE individuals who are not contingent in this domain (Park & Crocker,
2008). As compared to individuals with high non-contingent SE (i.e. secure SE)
and with LSE (whether contingent or non-contingent), individuals with high
contingent SE (i.e. fragile SE) experience more negative feelings and a greater
decline in state self-esteem after infidelity of their intimate partner or after a
professional failure (Zeigler-Hill et al., 2011).
Discrepant HSE is also related to self-enhancement strategies and
defensive behaviors. Jordan and colleagues (Jordan, Spencer, Zanna, HoshinoBrowne, & Correll, 2003) found that individuals with discrepant HSE are
more prone to in-group bias and rationalization of their decisions as a form
of reduction of cognitive dissonance, than individuals with congruent HSE.
Also, incongruent HSE was related to unrealistic optimism and self-idealization
(Bosson, Brown, Zeigler-Hill, & Swann, 2003), lower tendency to forgive and
more readiness to revenge for transgression (Eaton, Struthers, Shomrony, &
Santelli, 2007), and more self-handicapping in situations when success was
diagnostic of high abilities (i.e. when solving a very difficult task) (Lupien,
Seery, & Almonte, 2010).
This study aims to explore convergent validity of two aspects of fragile
HSE: contingent HSE and incongruent HSE (a combination of high ESE
and low ISE). To measure ISE, two best validated techniques, SE-IAT and
NLP (Bosson, Swann Jr, & Pennebaker, 2000), were used. Previous research
has shown that these two measures do not correlate with each other (Bosson
Bojana Bodroža
377
et al., 2000; Buhrmester, Blanton, & Swann Jr, 2011; Rudolph, SchröderAbé, Schütz, Gregg, & Sedikides, 2008). The lack of correlation has several
possible explanations regarding the measurement and the construct of ISE: a)
ISE as a construct is non-existent, b) only one of these instruments measures
ISE, while the other does not, c) none of the instruments measures ISE, d)
ISE is heterogeneous and different instruments measure its different aspects
(Bosson et al., 2000; Rudolph et al., 2008). Buhrmester (Buhrmester et al.,
2011) also suggested that SE-IAT measures implicit affect towards the self,
while NLP measures implicit egotism. Despite somewhat discouraging evidence
of its construct validity, research on ISE is flourishing. The continuing use of
measures of ISE was justified by their ability to capture the pronounced positive
bias towards the self, their satisfactory psychometrical characteristics, and their
ability to predict spontaneous, uncontrolled and non-verbal behaviors, as well as
others’ ratings of the individual’s behavior, even beyond the variance predicted
by explicit measures (Asendorpf, Banse, & Mücke, 2002; Conner & Barrett,
2005; Robinson & Meier, 2005; Rudolph, Schröder-Abé, Riketta, & Schütz,
2010; Spalding & Hardin, 1999). This is the reason why both measures of ISE –
SE-IAT and NLP were included in this study.
Narcissism and fragility of self-esteem
The early theories on narcissism describe some of the features of fragile
self-esteem as an important aspect of narcissism. Kohut (1966) argues that
behind an overt grandiosity of narcissists, there is the covert feeling of insecurity
and inferiority. This description of the dynamics of narcissism unequivocally
resembles the interplay of the automatic and preconscious ISE and conscious
ESE. Contemporary research confirms these assumptions, showing that narcissists
have high explicit self-esteem, which disguises low implicit self-esteem (Jordan
et al., 2003). Also, correlation between discrepant HSE and narcissism has been
found in a few studies (Jordan et al., 2003; Zeigler-Hill, 2006), but the results
seem to be inconsistent (for review see Bosson et al., 2008). Other aspects of
fragile self-esteem, such as the need for continual promotion and confirmation
of one’s self-worth, are also relevant for narcissism. Research has shown that
the subtype of narcissism called vulnerable narcissism is related to global
contingency of self-worth, whereas grandiose subtype is more domain specific.
Grandiose narcissists are more contingent upon being better than others, but less
on others’ approval (Zeigler-Hill, Clark, & Pickard, 2008).
Having in mind the theoretical significance of features of fragile selfesteem for the concept of narcissism and somewhat inconsistent research
findings regarding their relationship, we decided to address this topic in the
current research. The second goal of this study was to examine the relationship
between the two aspects of fragile self-esteem – incongruent and contingent
HSE, on the one hand, and narcissism on the other.
378
VALIDATION OF TWO CONCEPTUALIZATIONS OF FRAGILE SELF-ESTEEM:
CONTINGENT HIGH SELF-ESTEEM AND INCONGRUENT HIGH SELF-ESTEEM
Method
Sample
The sample consisted of 305 first and second year psychology students from the
University of Novi Sad and the University of Belgrade. This study was carried out in three
phases (described in detail in the Procedure section), which is why the number of participants
in each phase was different. The number of participants who answered every instrument is
presented in Table 1. Out of the total number, 269 participants took part in phases 1 and 2, 279
in phases 1 and 3, and 241 participants in phases 2 and 3. The average age of all respondents
included in the study was 20.10 years (SD=1.81, ranging from 18 to 30 years), with 82.1% of
the sample being female.
Procedure
The research was carried out as a part of a larger study. It was organized in three
phases in which the participants completed different questionnaires. In the first and the third
phase the participants answered the questionnaires in paper-and-pencil form, while in the
second phase they participated in computer-based testing in groups of up to 20 persons (every
respondent was seated individually at the computer). At the beginning of every phase, the
respondents were told that participation in the research was voluntary and all of them agreed
to participate. As for the instruments used, Rosenberg Self-Esteem Scale and NLP were used
in phase 1, SE-IAT in phase 2, and Contingent Self-Esteem Scale and Narcissist Personality
Inventory in phase 3.
Instruments
Contingent Self-Esteem Scale (Johnson & Blom, 2007) measures the extent to which
a person’s feeling of self-worth is based on two sources: personal competences and relations
with others. Competence based Self-Esteem (abbr. CBSE; 12 items, α=.78) refers to selfattitude contingent upon meeting personal standards of success, being perfect and avoiding
failures. Relation based Self-Esteem (abbr. RBSE; 14 items, α=.83) refers to the feeling of
self-worth which depends on the acceptance and love of others. The items are answered on
5-point Likert scale. The correlation between two subscales is r=.44.
The General Self-Esteem Scale (Opačić & Bodroža, in preparation) is an adapted and
extended version of Rosenberg Self-Esteem Scale (Rosenberg, 1965) which consists of 30
items. The scale includes equal number of positively (e.g. “I feel that I have a number of good
qualities.”) and negatively (e.g. “I feel I do not have much to be proud of.”) worded items.
The respondents choose the degree to which each statement refers to them on 5-point Likert
scale (1 – not at all, 5 – yes, completely). Internal consistency of the scale is α=.91.
Self-Esteem – Implicit Association Test (SE-IAT; Greenwald & Farnham, 2000;
Greenwald et al., 1998) measures the automaticity of association between the self and positive
attributes. The more automatized (i.e. faster) the association of the self with positive attributes
as compared to negative attributes, the higher the implicit self-esteem. The participants
were asked to categorize combinations of words from categories me / not me and pleasant/
unpleasant by pressing one of the two buttons as fast and as accurately as they could. As
suggested by the authors of the instrument (Greenwald & Farnham, 2000; Greenwald et al.,
1998), the participants were randomly assigned to one of the two versions of SE-IAT – one
with congruent blocks (me + pleasant) coming first, and the other with incongruent blocks
(me + unpleasant) coming first. Cohen’s d indicated strong SE-IAT effect (the difference
between mean latency for incongruent and congruent blocks; d=1.175).
Bojana Bodroža
379
To calculate the measure of SE-IAT effect, D algorithm, which controls for respondents’
mean latency, was used (Greenwald, Nosek, & Banaji, 2003)1. The analysis revealed the
difference in SE-IAT effect between the two versions (t(239)=4.892, p<.001, AMcongr=.331,
AMincongr=-.258). To annul the effect of the version, regression analysis was performed with
SE-IAT score as the criterion and version as the predictor variable, and the residual was saved
as the final SE-IAT measure. This measure was used in all analyses.
Name Letter Preference (NLP; Koole et al., 2001; Nuttin Jr, 1985, 1987) test is based
on the phenomenon of implicit egotism – a more pronounced preference for anything related to
oneself (in this case, personal name initials) over things not related to oneself (all other letters).
The participants rated 30 letters of Serbian alphabet on the 9-point scale (1 – don’t like it at all,
9 – like it very much). Following Nuttin Jr (1985) and Koole et al. (2001), the participants were
told that the research was concerned with people’s aesthetic judgments of simple stimuli like the
letters of the alphabet. The respondents were encouraged to answer intuitively, without much
thinking. The letters were presented in random order. NLP effect was calculated separately for
the name and the surname initial, using I-algorithm proposed by LeBel and Gawronski (2009).
The correlation between the two was only r=.38, so we decided to use first name and second
name initial preference separately in the analyses to avoid the possibility of type II error2.
Strong effect of implicit egotism was registered – preference for personal initials was stronger
than for non-initial letters (name initial: t(278)=20.24, p<.001, AMname=7.21, AMother.letters=4.89;
surname initial: t(278)=20.07, p<.001, AMsurname=7.30, AMother.letters=4.89).
Narcissistic Personality Inventory (NPI–16; Ames, Rose, & Anderson, 2006) is a
shortened version of NPI–40 (Raskin & Terry, 1988), which consists of 16 items. The authors
of this instrument conceptualize narcissism as a personality trait rather than a personality
disorder. The respondents are presented with pairs of statements and asked to decide which
one of these describes them better. Internal consistency of this instrument is somewhat below
the satisfactory level (α=.66).
Results
Descriptive statistics for all variables are presented in Table 1. In accordance
with the previous research that showed positive bias of explicit self-regard among
non-clinical population, i.e. a tendency of psychologically healthy individuals to
show rather high levels of self-esteem (Baumeister, Tice, & Hutton, 1989; Schimel,
Arndt, Pyszczynski, & Greenberg, 2001; Taylor & Brown, 1988), the measure of
ESE has negative standardized skewness. Negative standardized skewness was
also registered for all ISE measures. For the purpose of planned statistical analyses,
the distributions of these variables were normalized using Blom’s formula.
1 Initial analyses revealed low split-half reliability (.44). Since low reliability of measures can
lead to type II error in planned analyses (hierarhical regressions with interaction effects),
we decided to lower the error variance by calculating SE-IAT effect as was suggested by
Cunningham, Preacher, & Banaji (2001) and Nosek & Smyth (2007). Sixty trials from
practice blocks and critical blocks were divided into 4 sets of 15 trials. Four D measures
were calculated for every set and then principal components analysis was performed on
these four variables. The participants’ scores on the first principal component (48% of
variance explained) were taken as the SE-IAT measure.
2 Some research suggests that it is possible that NLPname and NLPsurname tap into somewhat
different constructs (Stieger & Burger, 2010). Namely, peoples’ first name initial is more
closely linked to their personal self-esteem, whereas surname initial is more closely linked
to collective self-esteem (Pelham, Mirenberg, & Jones, 2002).
380
VALIDATION OF TWO CONCEPTUALIZATIONS OF FRAGILE SELF-ESTEEM:
CONTINGENT HIGH SELF-ESTEEM AND INCONGRUENT HIGH SELF-ESTEEM
Table 1
Descriptive statistics of all variables
N
Min
Max
AM
SD
Standard.
Skewness
Standard.
Kurtosis
ESE
305
53
150
114.57
15.26
–4.91
CBSE
279
13
51
32.00
6.73
1.00
.13
RBSE
279
18
68
40.34
8.49
-.06
.38
SE-IAT
290
–3.60
2.68
0.00
1.00
–3.94
2.12
NLPname
279
–3.09
7.79
3.44
2.08
–5.19
.60
NLPsurname
279
–4.65
3.12
2.37
–6.05
1.43
Narcissism
279
6.32
3.13
2.36
–1.15
7.64
0
16
4.35
Legend: ESE –explicit self-esteem; CBSE – Competence based Self-Esteem; RBSE – Relation based
Self-Esteem; SE-IAT – Self-Esteem – Implicit Association Test; NLPname – Name Letter Preference
effect for name initial; NLPsurname – Name Letter Preference effect for surname initial.
Correlations between two measures of contingent self-esteem (CBSE and
RBSE), three measures of implicit self-esteem (SE-IAT, NLPname, NLPsurname), and
explicit self-esteem (ESE) were first analyzed. The results show that none of the
ISE measures correlates significantly with contingency of self-worth (Table 2).
Out of the three measures of ISE, only NLPname correlates positively with explicit
self-esteem. Therefore, the positivity of self-regard is accompanied by a stronger
preference for the first name initial as compared to all other letters. Both CBSE
and RBSE correlate negatively with ESE, suggesting that the more one’s feeling
of self-worth is based on meeting certain standards of success or being loved
and accepted by others, the lower their explicit self-esteem is.
Table 2
Intercorrelations between variables
1
2
3
4
5
6
1. CBSE
2. RBSE
.440**
3. SE-IAT
-.054
.005
4. NLPname
.000
-.097
5. NLPsurname
.055
.057
6. ESE
7. Narcissism
–.324
.007
**
.007
**
-.449
-.236
**
.066
.384**
.071
.156*
-.054
.195
**
-.007
-.006
.402**
** Correlation is significant at the p≤.01 level (2-tailed);
Legend: CBSE – Competence based Self-Esteem; RBSE – Relation based Self-Esteem; SE-IAT –
Implicit Association Test; NLPname – Name Letter Preference effect for name initial; NLPsurname – Name
Letter Preference effect for surname initial; ESE –explicit self-esteem.
The next step was to examine the relationship between CSE and a
combination of ISE and ESE. Separate hierarchical regression analyses were
381
Bojana Bodroža
carried out, with two measures of contingent self-esteem as criterion variables
and measures of ISE, ESE and their interaction term – ISExESE as predictor
variables (as suggested by Aiken & West (1991), predictors were first centered
and then the interaction term was calculated). Since three ISE measures
showed no correlation or low correlation with each other, they were entered
simultaneously in the regression analysis. Simple effects of predictors were
entered in Step 1 of regression analysis and the interaction terms of ESE with
measures of ISE in Step 2. The results of these analyses are presented in Table 3.
In both analyses only the first model, with simple effect of ISE and ESE
as predictors, was statistically significant, and the only significant predictor of
CBSE and RBSE was ESE. Adding the interaction term of ESE and all three
measures of ISE in Step 2 did not improve the predictive power of regression
models. Thus, regardless of the measure of ISE used, analyses do not support
the idea that a combination of high ESE and low ISE, i.e. discrepant high selfesteem is related to contingent self-esteem.
Table 3
Hierarchical regression models predicting subdimensions of contingent self-esteem
Criterions Predictors
Step 1
ESE
SE-IAT
NLPname
NLPsurname
Step 2
SE-IAT x ESE
NLPname x ESE
NLPsurname x ESE
ΔR2
.120***
CBSE
β
r
-.347***
-.008
.066
.023
-.338
-.045
.020
.053
-.006
.032
-.107
-.002
.043
-.110
.010
ΔR2
.199***
RBSE
β
r
-.441***
.044
-.048
.119
-.445
-.001
-.073
.111
-.029
.078
-.033
-.031
.114
.016
.006
*** p≤.001
Legend: CBSE – Competence based Self-Esteem; RBSE – Relation based Self-Esteem; ESE – explicit
self-esteem; SE-IAT – Implicit Association Test; NLPname – Name Letter Preference effect for name
initial; NLPsurname – Name Letter Preference effect for surname initial.
Nevertheless, the analyses reported above did not directly test the
hypothesis that the combination of high ESE and high CSE (i.e. contingent
high SE) was related to the combination of high ESE and low ISE. Since
there is no analysis that makes it possible to test the relationship between
two interactions (ESE x CSE and ESE x ISE), the decision was made to use
a less optimal approach of carrying out analyses on the subsample of HSE
individuals. The participants were ranked by their score on ESE and only
the upper 50% of the sample was selected (n=119). All analyses are repeated
on the subsample of HSE individuals. The correlations between CBSE and
RBSE, on one hand, and three measures of ISE, on the other hand, are all nonsignificant (rs≤|±.108|, ps>.05).
382
VALIDATION OF TWO CONCEPTUALIZATIONS OF FRAGILE SELF-ESTEEM:
CONTINGENT HIGH SELF-ESTEEM AND INCONGRUENT HIGH SELF-ESTEEM
The same hierarchical regression analyses were then carried out on the
subsample of HSE participants. The simple effects of predictors (ESE, SE-IAT,
NLPname, and NLPsurname) were entered in the first step, while the interaction
terms of ESE and all measures of ISE were entered in the second step of
regression analyses predicting CBSE and RBSE. When CBSE was used as
a criterion variable, neither the first nor the second regression model was
significant (first model: R2=.046, p>.05, second model ΔR2=.018, p>.05).
When RBSE was used as the criterion, only the first model was statistically
significant (R2=.110, p=.01), while the interaction terms entered in the second
step did not improve the predictive power of the regression model (ΔR2=.026,
p>.05). Consistent with the results of the analyses performed on the whole
sample, ESE is the only significant predictor of RBSE on the subsample of
HSE individuals (β=-.328, p<.001).
To sum up, discrepant SE was not related to contingent self-esteem
when all participants (LSE and HSE) were included in the analyses. But more
importantly and contrary to the theoretical expectations, discrepant SE did not
converge with contingent SE of HSE individuals. Thus, this research does not
provide evidence of convergent validity of the two aspects of fragile self-esteem
– discrepant self esteem and high contingent self-esteem.
Relationship between aspects of fragile self-esteem and narcissism
The second goal was to examine the relationship between aspects of fragile
SE and narcissism. Simple Pearson correlation (Table 2) shows that RBSE is
negatively correlated to narcissism, saying that narcissists do not have a need to
be accepted in order to feel good about themselves. They might even feel that
others’ non-acceptance is a sign of envy, which proves narcissists’ superiority.
Narcissists also have higher ESE. Out of three measures of implicit self-esteem,
only NLPname correlates with narcissism. So, narcissists manifest implicit egotism
considering their name initial, but not their surname initial.
Finally, hierarchical regression analyses were carried out to examine if
combinations of ISE and ESE, as well as CSE (i.e. subdimensions of CBSE and
RBSE) and ESE predict narcissism. Again, all three measures of ISE – SE-IAT,
NLPname, and NLPsurname were entered in the same regression analysis. Simple
effects of predictors were entered in the Step 1, and their interaction terms with
ESE in the Step 2 (predictors were first z-transformed).
In regression analyses where measures of ISE (SE-IAT, NLPname, and
NLPsurname), ESE, and the interaction term of ESE with ISE measures were used
as predictors of narcissism, only the first model with simple predictor effects
was significant (Table 4). In accordance with the Pearson correlations reported
in Table 2, only ESE and NLPname predicted narcissism. Models with interaction
terms did not reach the significance level, which leads to the conclusion that the
participants with congruent and incongruent levels of self-esteem have similar
levels of narcissism.
Bojana Bodroža
383
Table 4
Hierarchical regression model predicting narcissism based on explicit and implicit selfesteem and their interactions
Predictors
Step 1
ESE
SE-IAT
NLPname
NLPsurname
Step 2
SE-IAT x ESE
NLPname x ESE
NLPsurname x ESE
ΔR2
.173***
β
r
.365***
-.102
.141*
-.078
.402
-.055
.195
-.006
.011
.050
-.078
-.009
-.038
-.055
.007
*p<.05; ***p<.001
Legend: ESE – explicit self-esteem; SE-IAT – Implicit Association Test; NLPname – Name Letter
Preference effect for name initial; NLPsurname – Name Letter Preference effect for surname initial.
Further, narcissism was regressed on ESE, CBSE and RBSE in Step 1
and the interaction term of ESE with both dimensions of contingent self-esteem
in Step 2. The first model was statistically significant, while the model with
interaction terms was marginally significant (Table 5). Except for ESE and RBSE
that predict narcissism in the same way as indicated by Pearson correlations
(Table 2), CBSE is also a significant predictor of narcissism. The suppression
effect of CBSE – the discrepancy of non-significant zero-order correlation and
significant β coefficient, was registered. It suggests that, when ESE and RBSE
were leveled across participants, narcissists’ feeling of self-worth was also
contingent upon avoiding failures and meeting high personal standards.
In the second regression model, the interaction effect of RBSE and ESE
was not significant, but the interaction effect of CBSE and ESE was. The simple
slope test found that the slope of line representing the association between
CBSE and narcissism is significant for the participants with low ESE (β=.29,
p<.001), and marginally significant for HSE individuals (β=.14, p=.067) (the
levels of ESE are fixed at ±1SD). The predicted values for interaction are
presented in Figure 1. The individuals with low self-esteem who are highly
contingent upon proving their competences show significantly higher levels of
narcissism than those with low, but non-contingent self-esteem. Also, high selfesteem individuals with the feeling of self-worth that is contingent upon being
successful and competent have higher levels of narcissism than non-contingent
high self-esteem individuals. In accordance with the theoretical expectations, the
group of participants with high contingent self-esteem shows the highest level
of narcissism of all groups. Nevertheless, it should be noted that the registered
interaction effect is also the suppressor effect (zero-order correlation for the
interaction effect is non-significant), which means that only after participants
have been leveled by ESE, CBSE, and RBSE, can the combination of ESE and
CBSE additionally predict narcissism.
384
VALIDATION OF TWO CONCEPTUALIZATIONS OF FRAGILE SELF-ESTEEM:
CONTINGENT HIGH SELF-ESTEEM AND INCONGRUENT HIGH SELF-ESTEEM
Table 5
Hierarchical regression model predicting narcissism based on dimensions of contingent
self-esteem, explicit self-esteem and their interactions
Predictors
Step 1
ESE
CBSE
RBSE
Step 2
CBSE x ESE
RBSE x ESE
ΔR2
.209***
β
r
.416***
.268***
-.139*
.402
.007
-.236
-.161*
.055
-.025
-.003
.018†
*p<.05; ***p<.001; †p=.054.
Legend: CBSE – Competence Based Self-Esteem; RBSE – Relation Based Self-Esteem; ESE – explicit
self-esteem.
Figure 1. Narcissism as a function of explicit self-esteem and competence
based self-esteem
Discussion
This study set out to find evidence of convergent validity of two aspects
of fragile self-esteem: contingent high self-esteem and incongruent high selfesteem (i.e. a combination of high explicit and low implicit self-esteem).
Contrary to theoretical expectations, the results have shown no convergence
between these aspects of fragile self-esteem. That means that incongruent
high self-esteem is not necessarily dependent on the acceptance of others and
proving one’s own competences at the same time. These findings imply that
the concept of fragile self-esteem, in the sense it is understood now, has several
Bojana Bodroža
385
different aspects. Previous research found only limited evidence of convergence
of some aspects of fragile self-esteem. Researchers found modest correlations
between contingent, implicit, and unstable self-esteem (Crocker & Wolfe, 2001;
Kernis et al., 1993; Kernis & Paradise, 2002; Kernis et al., 2008), but significant
correlations between contingent and implicit self-esteem have not been found
in the present study. More importantly, in the previous research interactions of
level of self-esteem with its contingency, stability or congruency with implicit
self-esteem were not properly tested. To our knowledge, this research is the first
to test the relationship of high contingent and high incongruent self-esteem. The
results obtained are not promising from the perspective of the construct validity
of fragile self-esteem.
The results of this study call into question the concept of fragile selfesteem. Is it justifiable to consider two completely unrelated patterns of cognition
as aspects of the same psychological phenomenon? The empirical evidence
presented here does not provide support for such an interpretation. Even if we
accept the construct of fragile self-esteem to be valid, but highly diverse, we
should be careful when interpreting the results of the studies that used different
aspects of fragile self-esteem to predict certain behaviors.
Defensive behavior after ego threat is the most frequently examined
consequence of fragile self-esteem. Research has shown that all aspects of
fragile high self-esteem: stability, contingency and incongruency can predict
self-enhancing and defensive reactions to ego threats (Bosson et al., 2003; Eaton
et al., 2007; Jordan et al., 2003; Lupien et al., 2010; Kernis et al., 1993; Park &
Crocker, 2008; Zeigler-Hill et al., 2011). Nevertheless, it should be questioned
whether that is enough to consider them the aspects of the same psychological
phenomenon.
What makes the comparison of results even more complicated is that
the defensive reactions to ego threats measured in previous research were very
diverse in nature. Some of them were cognitive, some affective, and some
behavioral. Also, they differed in the way of measurement and included selfreported, indirect behavioral, as well as other-reported measures (Robinson &
Meier, 2005; Rudolph et al., 2010). The heterogeneity of self-enhancing and
defensive reactions measured in previous research makes it difficult to compare
the findings. It is possible that different aspects of fragile self-esteem predict
defensive reactions of different kinds. For example, Buhrmeister (Buhrmester et
al., 2011) concluded that SE-IAT measures implicit affect. As a consequence, this
measure of implicit self-esteem may be better at predicting affective reactions to
ego threats than cognitive ones. The research in which all three aspects of fragile
self-esteem would be used to predict the set of diverse defensive reactions
would be a very useful step in the validation of this construct and differentiation
between its aspects.
We cannot disregard important limitations of the analyses presented in this
paper. To our knowledge, there is no statistical analysis that makes it possible
to directly test the relationship between two interaction effects − interaction of
386
VALIDATION OF TWO CONCEPTUALIZATIONS OF FRAGILE SELF-ESTEEM:
CONTINGENT HIGH SELF-ESTEEM AND INCONGRUENT HIGH SELF-ESTEEM
contingency and the level of explicit self-esteem, and interaction of the level of
explicit and implicit self-esteem. Having that in mind, as well as the fact that
our theoretical hypotheses were related only to high self-esteem individuals, we
chose to perform additional analyses only on this subsample. Such a procedure
led not only to the restriction of range of explicit self-esteem, but also to lower
statistical power of statistical tests. Future research should seek for alternative
statistical approaches, which would better suit this kind of data.
Our second goal was to determine if fragile self-esteem characterizes
individuals with narcissistic personality characteristics. Students with highest
narcissism were found to have higher levels of explicit self-esteem than nonnarcissistic students, but to care less about the acceptance of others. More
interestingly, individuals with high levels of explicit self-esteem who feel
worthy only if they meet high standards of excellence or if they achieve success,
manifested the highest levels of narcissist personality trait. The results are
largely consistent with those of Zeigler-Hill who found that individuals with
grandiose narcissism have a strong need to be better than others, but do not care
about others’ acceptance and others’ opinions about themselves (Zeigler-Hill et
al., 2008). However, our findings suggest that competence based contingency
is related to narcissism only if it is accompanied by high levels of explicit
self-esteem3. In other words, narcissist personality characteristics are highest
among high self-esteem students whose feeling of self-worth depends on being
successful. To conclude, the results show that high self-esteem of individuals
with narcissistic personality characteristics is secure in the sense that it is not
dependent upon others’ approval, but is fragile and sensitive to the proof of their
competences and success. However, it should be noted that this research was
conducted on the sample of students for whom competences and success might
be especially important. Replication of this finding on the sample representative
for general population is, thus, needed. Another limitation of the present study
is related to the fact that the obtained interaction effect is the suppression effect.
Therefore, further research on this topic is necessary.
This research showed that among low self-esteem individuals, those who
have higher levels of self-esteem contingency in the domain of competences
are characterized by higher levels of narcissist personality characteristics
(although not as high as in the group of high self-esteem individuals). That
means that individuals who have a strong need for confirmation of their personal
competences and who base their feeling of self-worth on personal success
develop less self-critical and less psychologically healthy attitude toward the
self, even if their general self-regard is negative.
Discrepant high self-esteem, i.e. a combination of high explicit and low
implicit self-esteem, was not found to be a feature of persons with narcissistic
characteristics. This result should not be considered surprising or unusual.
3 The questionnaire Narcissist Personality Inventory (NPI), which was used in this study,
measures grandiose, but not vulneraible type of narcissism (Zeigler-Hill et al., 2008).
Moreover, Zeigler-Hill and colleagues used NPI for measuring grandiose narcissism,
which makes our results directly comparable.
Bojana Bodroža
387
Bosson and colleagues (Bosson et al., 2008) reviewed a large body of research
on this topic and concluded that the relationship between narcissism and
incongruent self-esteem is highly inconsistent. However, our study indicated
that one out of three measures of implicit self-esteem was related to narcissistic
personality trait. Inflated preference for one’s name initial, but not for the
surname initial, was related to higher narcissism. It seems that individuals with
the highest narcissism self-enhance through identification with their name initial
as the symbol of themselves as individuals, but not with their surname initial,
although it symbolically represents the group they belong to – their family.
This is consistent with the idea that the fist name initial is closely related to
personal self-esteem, whereas the last name initial is more closely related to
collective self-esteem (Pelham et al., 2002). However, it should be noted that
some research failed to find the relationship between narcissism and both NLP
and SE-IAT measures of implicit self-esteem (Zeigler-Hill, 2006).
It can be concluded that there is no clear evidence that fragile self-esteem
is the feature of narcissistic personalities. Our research offers limited evidence
that narcissistic personality characteristics are characterized by high self-esteem
contingent upon success, but not by discrepant high self-esteem.
Although this research did not focus on the validity of implicit self-esteem
measures, it is worth mentioning that IAT did not correlate to any other selfreported measure included in this study. Such results usually imply doubtful
validity of the instrument. Nevertheless, a number of previous studies showed
that IAT predicts spontaneous, uncontrolled, non-verbal, and other-reported
behaviors better than self-reported measures (Asendorpf et al., 2002; Conner &
Barrett, 2005; Robinson & Meier, 2005; Rudolph et al., 2010).
On the other hand, Name-Letter Preference test (i.e. only NLP for first
name initial), which is a self-report measure, correlated significantly but weakly
with explicit self-esteem and narcissism. The meta-analysis on the relationship
between NLP and explicit self-esteem showed comparable results (Krizan &
Suls, 2008), although the NLP measures included in this study averaged for the
name letter and surname letter. Although the participants in the present study
expressed implicit egotism for both name and surname initials, only pronounced
preference for name initial was related to explicit self-esteem and narcissistic
characteristics. In accordance with the conclusion of Pelham (Pelham et al.,
2002), it might suggest that, while the surname initial is a symbol of the group
participants belong to – their family, the name initial is a personal symbol and,
thus, relates more closely to the feeling of personal worth.
Conclusion
The present study does not offer proof of convergence between two aspects
of fragile self-esteem: contingent high self-esteem and discrepant high selfesteem. Further research on the validity of this construct should also include the
third aspect of self-esteem fragility – stability of self-esteem. Previous research
388
VALIDATION OF TWO CONCEPTUALIZATIONS OF FRAGILE SELF-ESTEEM:
CONTINGENT HIGH SELF-ESTEEM AND INCONGRUENT HIGH SELF-ESTEEM
on fragile self-esteem as predictor of defensive reactions to ego threats included
a wide range of different reactions (affective, cognitive, and behavioral) and the
ways of their measurement (direct, indirect, self-reported, other-reported, etc).
This field would benefit from the research in which all three aspects of fragile
self-esteem would be compared as the predictors of the same set of defensive
reactions. Future research on narcissism and fragility of self-esteem should also
focus on other aspects of fragile self-esteem (e.g. unstable high self-esteem) or,
even better, to all three aspects together.
References
Aiken, L. S., & West, S. G. (1991). Multiple Regression: Testing and Interpreting Interactions.
Newbury Park: Sage.
Ames, D. R., Rose, P., & Anderson, C. P. (2006). The NPI–16 as a short measure of narcissism.
Journal of Research in Personality, 40(4), 440–450. doi:10.1016/j.jrp.2005.03.002
Asendorpf, J. B., Banse, R., & Mücke, D. (2002). Double dissociation between implicit and
explicit personality self-concept: The case of shy behavior. Journal of Personality and
Social Psychology, 83(2), 380–393. doi:10.1037/0022–3514.83.2.380
Battle, J., Jarratt, L., Smit, S., & Precht, D. A. N. (1988). Relations Among Self-Esteem,
Depression, and Anxiety of Children. Psychological Reports, 62(3), 999–1005.
doi:10.2466/pr0.1988.62.3.999
Baumeister, R. F. (1998). The Self. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The
Handbook of Social Psychology (Vol. 1, pp. 680–740). New York: McGraw-Hill.
Baumeister, R. F., Smart, L., & Boden, J. M. (1996). Relation of Threatened Egotism to
Violence and Aggression: The Dark Side of High Self-Esteem. Psychological Review,
103(1), 5–33.
Baumeister, R. F., Tice, D. M., & Hutton, D. G. (1989). Self-Presentational Motivations
and Personality Differences in Self-Esteem. Journal of Personality, 57(3), 547–579.
doi:10.1111/j.1467–6494.1989.tb02384.x
Bosson, J. K., Brown, R. P., Zeigler-Hill, V., & Swann, W. B. (2003). Self-Enhancement
Tendencies Among People With High Explicit Self-Esteem: The Moderating Role of
Implicit Self-Esteem. Self & Identity, 2(3), 169–187.
Bosson, J. K., Lakey, C. E., Campbell, W. K., Zeigler-Hill, V., Jordan, C. H., & Kernis, M.
H. (2008). Untangling the Links between Narcissism and Self-esteem: A Theoretical
and Empirical Review. Social and Personality Psychology Compass, 2(3), 1415–1439.
doi:10.1111/j.1751–9004.2008.00089.x
Bosson, J. K., Swann Jr, W. B., & Pennebaker, J. W. (2000). Stalking the perfect measure of
implicit self-esteem: The blind men and the elephant revisited? Journal of Personality and
Social Psychology, 79(4), 631–643. doi:10.1037/0022–3514.79.4.631
Brown, J. D. (2012). Understanding the Better Than Average Effect. Personality and Social
Psychology Bulletin, 38(2), 209–219. doi:10.1177/0146167211432763
Buhrmester, M. D., Blanton, H., & Swann Jr, W. B. (2011). Implicit self-esteem: Nature,
measurement, and a new way forward. Journal of Personality and Social Psychology,
100(2), 365–385. doi:10.1037/a0021341
Campbell, J. D., Chew, B., & Scratchley, L. S. (1991). Cognitive and Emotional Reactions to
Daily Events: The Effects of Self-Esteem and Self-Complexity. Journal of Personality,
59(3), 473–505. doi:10.1111/1467–6494.ep9110141809
Colvin, C. R., Block, J., & Funder, D. C. (1995). Overly positive self-evaluations and
personality: Negative implications for mental health. Journal of Personality and Social
Psychology, 68(6), 1152–1162. doi:10.1037/0022–3514.68.6.1152
Bojana Bodroža
389
Conner, T., & Barrett, L. F. (2005). Implicit Self-Attitudes Predict Spontaneous Affect in
Daily Life. Emotion, 5(4), 476–488. doi:10.1037/1528–3542.5.4.476
Crocker, J. (2002). Contingencies of Self-Worth: Implications for SelfRegulation and Psychological Vulnerability. Self & Identity, 1(2), 143–149.
doi:10.1080/152988602317319320
Crocker, J., & Wolfe, C. T. (2001). Contingencies of Self-Worth. Psychological Review,
108(3), 593–623.
Cunningham, W. A., Preacher, K. J., & Banaji, M. R. (2001). Implicit Attitude Measures:
Consistency, Stability, and Convergent Validity. Psychological Science, 12(2), 163–170.
Diener, E. (1984). Subjective well-being. Psychological Bulletin, 95(3), 542–575.
doi:10.1037/0033–2909.95.3.542
Eaton, J., Struthers, C. W., Shomrony, A., & Santelli, A. G. (2007). When apologies fail: The
moderating effect of implicit and explicit self-esteem on apology and forgiveness. Self &
Identity, 6(2/3), 209–222. doi:10.1080/15298860601118819
Greenwald, A. G., & Banaji, M. R. (1995). Implicit social cognition: Attitudes, self-esteem,
and stereotypes. Psychological Review, 102(1), 4–27. doi:10.1037/0033–295x.102.1.4
Greenwald, A. G., & Farnham, S. D. (2000). Using the Implicit Association Test to measure
self-esteem and self-concept. Journal of Personality and Social Psychology, 79(6), 1022–
1038. doi:10.1037/0022–3514.79.6.1022
Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. K. (1998). Measuring individual
differences in implicit cognition: The implicit association test. Journal of Personality and
Social Psychology, 74(6), 1464–1480. doi:10.1037/0022–3514.74.6.1464
Greenwald, A. G., Nosek, B. A., & Banaji, M. R. (2003). Understanding and using the Implicit
Association Test: I. An improved scoring algorithm. Journal of Personality and Social
Psychology, 85(2), 197–216. doi:10.1037/0022–3514.85.2.197
Janis, I. L. (1954). Personality Correlates of Susceptibility To Persuasion. Journal of
Personality, 22(4), 504. doi:10.1111/1467–6494.ep8930397
Johnson, M., & Blom, V. (2007). Development and Validation of Two Measures of Contingent
Self-Esteem. Individual Differences Research, 5(4), 300–328.
Jordan, C. H., Spencer, S. J., Zanna, M. P., Hoshino-Browne, E., & Correll, J. (2003). Secure
and Defensive High Self-Esteem. Journal of Personality and Social Psychology, 85(5),
969–978. doi:10.1037/0022–3514.85.5.969
Kernis, M. H. (2003). Toward a Conceptualization of Optimal Self-Esteem. Psychological
Inquiry, 14(1), 1–26.
Kernis, M. H., Abend, T. A., Goldman, B. M., Shrira, I., Paradise, A. N., & Hampton, C.
(2005). Self-serving responses arising from discrepancies between explicit and implicit
self-esteem. Self & Identity, 4(4), 311–330. doi:10.1080/15298860500146028
Kernis, M. H., Cornell, D. P., Sun, C.-R., Berry, A., & Harlow, T. (1993). There’s more to
self-esteem than whether it is high or low: The importance of stability of self-esteem.
Journal of Personality and Social Psychology, 65(6), 1190–1204. doi:10.1037/0022–
3514.65.6.1190
Kernis, M. H., Lakey, C. E., & Heppner, W. L. (2008). Secure Versus Fragile High Self-Esteem
as a Predictor of Verbal Defensiveness: Converging Findings Across Three Different
Markers. Journal of Personality, 76(3), 477–512. doi:10.1111/j.1467–6494.2008.00493.x
Kernis, M. H., & Paradise, A. N. (2002). Distinguishing between secure and fragile forms
of high self-esteem. In E. L. Deci & R. M. Ryan (Eds.), Handbook of self-determination
research (pp. 339–360). Rochester, NY: University Of Rochester Press.
Kohut, H. (1966). Forms and Transformations of Narcissism. Journal of the American
Psychoanalytic Association, 14(2), 243–272. doi:10.1177/000306516601400201
Koole, S. L., Dijksterhuis, A., & van Knippenberg, A. (2001). What’s in a name: Implicit
self-esteem and the automatic self. Journal of Personality and Social Psychology, 80(4),
669–685. doi:10.1037/0022–3514.80.4.669
390
VALIDATION OF TWO CONCEPTUALIZATIONS OF FRAGILE SELF-ESTEEM:
CONTINGENT HIGH SELF-ESTEEM AND INCONGRUENT HIGH SELF-ESTEEM
Krizan, Z., & Suls, J. (2008). Are implicit and explicit measures of self-esteem related? A
meta-analysis for the Name-Letter Test. Personality and Individual Differences, 44(2),
521–531.
Leary, M. R., & Kowalski, R. M. (1995). Social Anxiety. New York: Guilford Press.
LeBel, E. P., & Gawronski, B. (2009). How to find what’s in a name: Scrutinizing the
optimality of five scoring algorithms for the name-letter task. European Journal of
Personality, 23(2), 85–106. doi:10.1002/per.705
Lupien, S. P., Seery, M. D., & Almonte, J. L. (2010). Discrepant and congruent high selfesteem: Behavioral self-handicapping as a preemptive defensive strategy. Journal of
Experimental Social Psychology, 46(6), 1105–1108.
McFarlin, D. B., Baumeister, R. F., & Blascovich, J. (1984). On Knowing When to Quit: Task
Failure, Self-Esteem, Advice, and Nonproductive Persistence. Journal of Personality,
52(2), 138–155.
Myers, D. G., & Diener, E. (1995). Who is happy? Psychological Science, 6(1), 10–19.
Nosek, B. A., & Smyth, F. L. (2007). A multitrait-multimethod validation of the Implicit
Association Test: Implicit and explicit attitudes are related but distinct constructs.
Experimental Psychology, 54(1), 14–29. doi:10.1027/1618–3169.54.1.14
Nuttin Jr, J. M. (1985). Narcissism beyond Gestalt and awareness: the name letter effect.
European Journal of Social Psychology, 15(3), 353–361.
Nuttin Jr, J. M. (1987). Affective consequences of mere ownership: The name letter effect
in twelve European languages. European Journal of Social Psychology, 17(4), 381–402.
Opačić, G., & Bodroža, B. (in preparation). Psychometric properties of the General SelfEsteem Scale.
Park, L. E., & Crocker, J. (2008). Contingencies of self-worth and responses to negative
interpersonal feedback. Self & Identity, 7(2), 184–203. doi:10.1080/15298860701398808
Paulhus, D. L. (1998). Interpersonal and intrapsychic adaptiveness of trait self-enhancement:
A mixed blessing? Journal of Personality and Social Psychology, 74(5), 1197–1208.
doi:10.1037/0022–3514.74.5.1197
Pelham, B. W., Mirenberg, M. C., & Jones, J. T. (2002). Why Susie sells seashells by the
seashore: Implicit egotism and major life decisions. Journal of Personality and Social
Psychology, 82(4), 469–487. doi:10.1037/0022–3514.82.4.469
Raskin, R., & Terry, H. (1988). A principal-components analysis of the Narcissistic Personality
Inventory and further evidence of its construct validity. Journal of Personality and Social
Psychology, 54(5), 890–902. doi:10.1037/0022–3514.54.5.890
Rawson, H. E. (1992). The Interrelationship of Measures of Manifest Anxiety, Self-Esteem,
Locus of Control, and Depression in Children with Behavior Problems. Journal of
Psychoeducational Assessment, 10(4), 319–329. doi:10.1177/073428299201000402
Robins, R. W., & Beer, J. S. (2001). Positive illusions about the self: Short-term benefits
and long-term costs. Journal of Personality and Social Psychology, 80(2), 340–352.
doi:10.1037/0022–3514.80.2.340
Robinson, M. D., & Meier, B. P. (2005). Rotten to the core: Neuroticism and implicit
evaluations of the self. Self & Identity, 4(4), 361–372. doi:10.1080/15298860500241852
Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton
University Press.
Rudolph, A., Schröder-Abé, M., Riketta, M., & Schütz, A. (2010). Easier when done than
said!: Implicit self-esteem predicts observed or spontaneous behavior, but not selfreported or controlled behavior. Zeitschrift für Psychologie/Journal of Psychology, 218(1),
12–19. doi:10.1027/0044–3409/a000003
Rudolph, A., Schröder-Abé, M., Schütz, A., Gregg, A. P., & Sedikides, C. (2008). Through
a glass, less darkly? Reassessing convergent and discriminant validity in measures of
implicit self-esteem. European Journal of Psychological Assessment, 24(4), 273–281.
doi:10.1027/1015–5759.24.4.273
Bojana Bodroža
391
Schimel, J., Arndt, J., Pyszczynski, T., & Greenberg, J. (2001). Being accepted for who we are:
Evidence that social validation of the intrinsic self reduces general defensiveness. Journal
of Personality and Social Psychology, 80(1), 35–52. doi:10.1037/0022–3514.80.1.35
Spalding, L. R., & Hardin, C. D. (1999). Unconscious unease and self-handicapping:
Behavioral consequences of individual differences in implicit and explicit self-esteem.
Psychological Science, 10(6), 535–539.
Stieger, S., & Burger, C. (2010). Implicit and Explicit Self-Esteem in the Context of
Internet Addiction. CyberPsychology, Behavior & Social Networking, 13(6), 681–688.
doi:10.1089/cyber.2009.0426
Taylor, S. E., & Brown, J. D. (1988). Illusion and well-being: A social psychological
perspective on mental health. Psychological Bulletin, 103(2), 193–210. doi:10.1037/0033–
2909.103.2.193
Tennen, H., & Affleck, G. (1993). The Puzzle of Self-Esteem: A Clinical Perspective. In R. F.
Baumeister (Ed.), The Self-Esteem: The Puzzle of Low Self-Regard (pp. 241–262). New
York: Plenum Press.
Vaux, A. (1988). Social and Emotional Loneliness: The Role of Social and Personal
Characteristics. Personality and Social Psychology Bulletin, 14(4), 722–734.
doi:10.1177/0146167288144007
Zeigler-Hill, V. (2006). Discrepancies Between Implicit and Explicit Self-Esteem: Implications
for Narcissism and Self-Esteem Instability. Journal of Personality, 74(1), 119–144.
doi:10.1111/j.1467–6494.2005.00371.x
Zeigler-Hill, V., Besser, A., & King, K. (2011). Contingent Self-Esteem and Anticipated
Reactions to Interpersonal Rejection and Achievement Failure. Journal of Social &
Clinical Psychology, 30(10), 1069–1096. doi:10.1521/jscp.2011.30.10.1069
Zeigler-Hill, V., Clark, C. B., & Pickard, J. D. (2008). Narcissistic Subtypes and Contingent
Self-Esteem: Do All Narcissists Base Their Self-Esteem on the Same Domains? Journal
of Personality, 76(4), 753–774. doi:10.1111/j.1467–6494.2008.00503.x